This is not data, but in my peer group there is a strong connection between those who dropped out and those who had self control issues with any of those.
Wasnt the article comparing pot users with non pot users? Presumably the control group included people who drink same play video games as well?
Probably pot users also drink an play video games, so they don't affect the difference in performance.
This is one of the problems with recreational drug studies. Its hard to find people who only do one drug and then people who live similar life styles to them but sober. I remember and MDMA study that looked at ravers in Utah because a large number of them were sober to study the effects of MDMA while accounting for the raver lifestyle of staying up for many hours dancing and often dehydrated.